tensorflow.jsのobjectdetectionで自分のmodelを使うとき

チュートリアルの

  cocoSsd.load().then(model => {
    // detect objects in the image.
    model.detect(img).then(predictions => {
      console.log('Predictions: ', predictions);
    });
  });

のloadの引数を

  cocoSsd.load({ base: 'lite_mobilenet_v2', modelUrl: './myOriginal/model.json' }).then(model => {
    // detect objects in the image.
    model.detect(img).then(predictions => {
      console.log('Predictions: ', predictions);
    });

とかにすればOK